The maximum concentration of hydrogen that can be commingled with or blended in natural gas has been a topic of discussion for many years. The reasons for blending hydrogen with natural gas are mainly twofold: to improve the renewable content of natural gas and to utilize the existing natural gas pipeline infrastructure for hydrogen distribution. This paper aims at addressing these issues by proposing an optimization model that considers blending hydrogen generated through a "power-to-gas" (PtG) technology pathway with natural gas. This paper develops a deterministic mixed-integer nonlinear programming network pressure-flowrate optimization model that captures the interaction between natural gas pipeline systems and the impact of hydrogen injections across multiple pipeline connections in the natural gas grid. A stochastic model is subsequently developed. The optimal quality/concentration of hydrogen in the natural gas pipelines and the optimal location of PtG units are realized while minimizing system-wide capital and operating costs. The design decisions obtained for the deterministic model differ from the stochastic model; the design decisions for the stochastic model varies across the different uncertainty realizations.